Project Summary Aggression is a fundamental social behavior. Though widespread, the stimuli that modulate aggression differ between species. Primates rely strongly on visual cues, while in rodents and insects olfactory stimuli are essential. Since mice and flies are the leading models of modern aggression studies, the mechanisms by which visual neural circuits modulate aggression remain largely unknown. Recent advances in genome sequencing, transgenic and viral technologies allow the application of genetically-encoded tools for tracing, monitoring and manipulating neural circuits to a wide variety of species. Here, we propose to develop an innovative model to uncover the circuitry underlying visually-evoked aggression. We will develop advanced tools and experimental setups to study the neuronal circuits and computations involved in visually-triggered aggression in Siamese fighting fish (Betta splendens). Betta have been selectively bred to be highly aggressive for hundreds of generations. They now have a fast, robust, and stereotyped aggression display in response to seeing another fish, and therefore present an ideal model for the proposed studies. We will first test whether visually-evoked aggression engages specific nodes in a circuit that receives visual input (Aim 1). We will identify these nodes in order to target them for analyses of neuronal activity. This will be achieved in a three step process. First we will anatomically trace vision-aggression circuits using anterograde tracers from the retina and retrograde tracers from the muscles used exclusively during aggression displays. We will then identify salient features of aggression-triggering visual cues, to develop visual stimuli that robustly elicit aggression and other stimuli that merely elicit visual attention but no aggression. Lastly, we will identify neurons active during aggression (using an antibody against pS6, a marker of neuronal activity) but not when fish see a control movie that elicits attention but not aggression. Together, this will enable us to identify the nodes involved in visually-evoked aggression, so we can target them for analyses of neuronal activity. We will then develop transgenic Betta to express genetically encoded calcium indicators in the brain to measure how visual stimuli that trigger aggression are represented and processed (Aim 2). Lastly, we will develop a head-fixed preparation to record neuronal activity in behaving fish, focusing first in the fish homolog of the mammalian amygdala (Aim 3). This large scale and cellular-level resolution approach will allow us to study the computations involved in the perception and discrimination of visual stimuli with varying aggression- inducing levels and to generate models that relate neuronal activity to behavioral output. The model developed in this application will present a unique and powerful platform to elucidate the neural circuitry underlying visually-evoked aggression, a project which we aspire to present as a subsequent TargetedBCP R01.